Our Work

"AI will change everything" is the laziest sentence in retail strategy. It tells you nothing about where to invest first, what to protect, or what to let go. AI agents will not disrupt all catego…

"AI will change everything" is the laziest sentence in retail strategy. It tells you nothing about where to invest first, what to protect, or what to let go. AI agents will not disrupt all categories simultaneously. The pace follows a gradient determined by one variable: how much of the purchase decision depends on objectively comparable specifications versus subjective, tactile, or experiential evaluation. Your category is not "safe" or "doomed." It is a portfolio of sub-segments sitting at different points on that gradient. The disaggregation is the strategy. I introduce the Category-Specific Vulnerability Gradient and five diagnostic variables practitioners can apply to their own categories.

A versioned collection of fourteen distinct AI-assisted research manuscripts on quantum information and quantum computation. Each manuscript is provided as a separate PDF, with its exact title, author…

A versioned collection of fourteen distinct AI-assisted research manuscripts on quantum information and quantum computation. Each manuscript is provided as a separate PDF, with its exact title, author, date, scope, website article and QIQC report identified in README.md and MANUSCRIPTS.json. Source and verification archives are preserved in sources-and-verification.zip. This is an independently authored collection, not an official publication of the QIQC catalog. The manuscripts and their verification qualifications are unchanged; archival publication does not establish novelty, external proof validation or catalog acceptance. Background references remain in each manuscript bibliography and are separate from the manuscripts reporting the results. Withdrawn partial results and research logs are excluded. Collection version: 1.0, 26 September 2026. Author: Yuxuan Zhang. Individual manuscript dates are preserved in the index and PDFs. The eleven-qubit manuscript has an existing individual DOI, https://doi.org/10.5281/zenodo.22968593, retained and linked as a constituent of this collection. Prefer that individual DOI when citing that manuscript alone. Research website: https://yuxuanzhang1995.github.io/agentic-research/

This record provides reproducibility materials supporting the manuscript “Actual-use verification and lineage-resolved fidelity in scientific AI”. The study evaluates scientific-AI fidelit…

This record provides reproducibility materials supporting the manuscript “Actual-use verification and lineage-resolved fidelity in scientific AI”. The study evaluates scientific-AI fidelity by distinguishing assigned scientific objects, realized scientific objects, lineage or pathway fidelity, terminal scientific fidelity and claim boundaries.  The archive contains detailed evidence registries, comparison specifications, provenance and audit records, structured source records, manuscript source materials, display assets, and code for assembling and rendering the supplied publication materials. It also preserves the available historical CoralTemp/Coral Reef Watch daily-classification source excerpt used in the documented evidence chain.  The released materials support inspection of the reported evidence, scientific-object assignments, realized-use verification, lineage-resolved comparisons and presentation-level reproduction. They do not constitute a complete executable end-to-end reproduction of all scientific analyses. Complete raw-input and terminal-reference payloads, full benchmark execution and scoring pipelines, original scientific figure-generation programs and the original pinned scientific execution environments are not included. Public upstream scientific products and their provenance are documented in the manuscript and archive.  The archive should therefore be interpreted as a provenance-rich reproducibility and audit package rather than as a general benchmark of scientific-AI accuracy or reliability.

Supplementary Material This repository contains supplementary Research material for the Master's Thesis: When Games Go "Woke": Mapping the DEI Controversy in the Gaming Culture War The file include th…

Supplementary Material This repository contains supplementary Research material for the Master's Thesis: When Games Go "Woke": Mapping the DEI Controversy in the Gaming Culture War The file include the main empirical Research corpus and network data used in the Analysis, the Gephi Project used to create the controversy map, and high-Resolution Images of the final controversy map.   Meurer_2026_Empirical Corpus.xlsx    This workbook contains:    - Empirical Corpus: The complete empirical corpus of 172 entries used for the      controversymapping and discursive Analysis    - Nodes: The complete node table used to construct the controversy map via Gephi.    - Edges: The complete edge table used to construct the controversy map via Gephi.   Meurer_2026_Controversy Map.gephi   The file contains the saved Gephi Project used to create the final controversy map. It   includes the imported network data as well as the layout and visualization settings used   for the final map.   Meurer_2026_Final Controversy Map.pdfMeurer_2026_Final Controversy Map.svg   Both files contain high-Resolution versions of the final controversy map presented in      the Master Thesis.

For a holomorphic function f, P´olya’s vector field f turns contour integrals into circula- tion and flux, and turns the poles of f into point charges. Applied to f= ζ′/&ze…

For a holomorphic function f, P´olya’s vector field f turns contour integrals into circula- tion and flux, and turns the poles of f into point charges. Applied to f= ζ′/ζ this produces the plane electrostatic field E= ∇log |ζ|, with a unit positive charge at every zero of ζ and a unit negative charge at the pole. We give a self-contained account of this dictionary and use it to read a number of classical statements about ζ as statements about a field: the functional equation as a wind of speed log(t/2π) blowing to the left, uniform across any fixed vertical strip left of the critical strip; the identity Re ζ′/ζ(1/2 + it) =−ϑ′(t) on the crit- ical line; Speiser’s theorem, which says that the Riemann Hypothesis is equivalent to every stagnation point of E in the critical strip lying to the right of the critical line; Lagarias’s positivity criterion, which says that it is equivalent to the horizontal component of the field of the completed function ξnever pointing toward the line; and the argument principle as the divergence theorem. On the prime side, the Dirichlet series of−ζ′/ζ makes the same field a superposition of rotors, one per prime power, and the smoothed explicit formula makes every zero a bump of exactly known shape in a smoothed prime sum. We show that differencing such sums across the cutoff cancels the term ζ′/ζ(s), leaving a remainder of size O(N−1/m), and that a deconvolution of the remaining bumps recovers, from the von Mangoldt function alone, the ordinates of the first fifteen zeros and of eleven zeros near height 1000 to within 10−3, and their real parts to within 2×10−4 and 5×10−3 respectively. Numerical experiments accompany each section: the zeros of ζ′up to height 300 against the Levinson–Montgomery sum rule, the field of a Dirichlet L-function, and the negative discrete moments of ζ′(ρ) over the first two thousand zeros against Gonek’s constant 3/π3. The paper is expository: its theorems are classical and attributed; the numerical arrangements are new to the author’s knowledge and are identified as such.

Jupyter notebooks for ME 50900, Intermediate Fluid Mechanics, at Purdue University. The notebooks cover tensors, kinematics, unidirectional flows, the Navier–Stokes equations, ideal flow, dimensional …

Jupyter notebooks for ME 50900, Intermediate Fluid Mechanics, at Purdue University. The notebooks cover tensors, kinematics, unidirectional flows, the Navier–Stokes equations, ideal flow, dimensional analysis, similarity solutions, boundary layers, Stokes flow, and lubrication theory, using Python with SymPy for symbolic calculations and Matplotlib for visualizations. Each notebook opens in Google Colab with nothing to install.

Analysis software at time of publication.

O desenvolvimento da linguagem infantil constitui um processo complexo, relacionado à aquisição de habilidades linguísticas, cognitivas, sociais e comunicativas. No Transto…

O desenvolvimento da linguagem infantil constitui um processo complexo, relacionado à aquisição de habilidades linguísticas, cognitivas, sociais e comunicativas. No Transtorno do Espectro Autista (TEA), podem ser observadas diferentes particularidades no desenvolvimento da comunicação, que variam significativamente entre as crianças e podem envolver aspectos da linguagem receptiva e expressiva, comunicação não verbal, interação social, pragmática, prosódia e uso funcional da linguagem. Nesse contexto, a atuação fonoaudiológica apresenta relevância na avaliação e no acompanhamento das habilidades comunicativas, considerando não apenas a produção verbal, mas também as diferentes formas utilizadas pela criança para expressar necessidades, interesses, emoções e intenções comunicativas. Este artigo tem como objetivo discutir as características da linguagem infantil no TEA e analisar as contribuições da Fonoaudiologia para o desenvolvimento da comunicação e da participação social dessas crianças. Trata-se de uma pesquisa de natureza bibliográfica, de abordagem qualitativa e caráter exploratório, fundamentada em produções científicas relacionadas ao desenvolvimento da linguagem, comunicação social, pragmática, ecolalia e intervenção fonoaudiológica no TEA. A literatura demonstra que a comunicação no autismo apresenta grande heterogeneidade, sendo necessária uma avaliação individualizada que considere as potencialidades, necessidades e formas particulares de comunicação de cada criança. Destaca-se, ainda, a importância de estratégias que favoreçam a comunicação funcional e a participação da família e dos diferentes contextos sociais no processo terapêutico.

Amyloid-related imaging abnormalities (ARIA) occur in 20-40% of patients treated with the amyloid-targeting drugs lecanemab and donanemab. While most cases of ARIA are asymptomatic, serious and fatal …

Amyloid-related imaging abnormalities (ARIA) occur in 20-40% of patients treated with the amyloid-targeting drugs lecanemab and donanemab. While most cases of ARIA are asymptomatic, serious and fatal cases occur. Therefore, ARIA prevention remains vital as these drugs are increasingly offered. Intensive blood pressure management (i.e., target systolic blood pressure of <120-130 mmHg) has been proposed to prevent ARIA. Though blood pressure control is key for neurovascular risk optimization and dementia prevention, the evidence supporting intensive blood pressure management for ARIA prevention is not yet robust enough to inform clinical practice. Extant observational studies do not demonstrate a consistent association between blood pressure and ARIA, are often unadjusted or inappropriately adjusted, and none directly evaluate intensive blood pressure treatment initiation. Meanwhile, the harms of intensive blood pressure treatment in older adults are immediate and well documented. Thus, we argue that current guideline-concordant management of hypertension should continue in older adults.